NZ793119A9 - Optimizing pallet location in a warehouse - Google Patents

Abstract

system for managing a plurality of pallets in a warehouse. The system includes: a plurality of storage racks having a plurality of rack openings; a database that is programmed to store pallet allocation data that associates pallet storage durations with the plurality of rack openings; and a computer system including one or more processors that are programmed to perform various operations. The various operations include: determining an expected storage duration of a target pallet in the warehouse; determining a height of the target pallet; calculating optimization values for the plurality of rack openings based on the expected storage duration and the height of the target pallet; identifying, among the plurality of rack openings, a target rack opening as a storage location for the target pallet; and transmitting information identifying the storage location to equipment for placement of the target pallet. The target rack opening has an optimization value that satisfies a preset requirement. Each of the optimization values includes a combination of a duration match value and a height match value for each of the plurality of rack openings. The duration match value for a rack opening with respect to a pallet represents proximity in duration value between the pallet storage duration associated with the rack opening and the expected storage duration of the pallet. The height match value for the rack opening with respect to the pallet represents proximity in measurement between a height of the rack opening and a height of the pallet. er system including one or more processors that are programmed to perform various operations. The various operations include: determining an expected storage duration of a target pallet in the warehouse; determining a height of the target pallet; calculating optimization values for the plurality of rack openings based on the expected storage duration and the height of the target pallet; identifying, among the plurality of rack openings, a target rack opening as a storage location for the target pallet; and transmitting information identifying the storage location to equipment for placement of the target pallet. The target rack opening has an optimization value that satisfies a preset requirement. Each of the optimization values includes a combination of a duration match value and a height match value for each of the plurality of rack openings. The duration match value for a rack opening with respect to a pallet represents proximity in duration value between the pallet storage duration associated with the rack opening and the expected storage duration of the pallet. The height match value for the rack opening with respect to the pallet represents proximity in measurement between a height of the rack opening and a height of the pallet.

Description

A system for ng a plurality of pallets in a warehouse. The system includes: a plurality of storage racks having a plurality of rack openings; a database that is programmed to store pallet allocation data that ates pallet storage durations with the plurality of rack openings; and a computer system including one or more processors that are programmed to perform various operations. The various operations include: determining an expected storage duration of a target pallet in the warehouse; determining a height of the target pallet; calculating optimization values for the plurality of rack openings based on the expected storage duration and the height of the target pallet; identifying, among the plurality of rack openings, a target rack opening as a storage location for the target pallet; and transmitting ation identifying the storage location to ent for placement of the target pallet. The target rack opening has an zation value that satisfies a preset requirement. Each of the optimization values includes a combination of a duration match value and a height match value for each of the plurality of rack openings. The duration match value for a rack opening with respect to a pallet represents proximity in duration value between the pallet e duration ated with the rack opening and the expected storage duration of the pallet. The height match value for the rack opening with respect to the pallet ents proximity in measurement between a height of the rack opening and a height of the pallet. 793119 A9 OPTHVJIZING PALLET ON INAWAREHOUSE TECHNICAL FIELD [0001[ This document describes devices, systems, and methods related to warehouse management by optimizing pallet locations in a use.

BACKGROUND [0002[ Warehouses include warehouse racks to store s of goods. Pallets are lly ?at transport structures that support goods in a stable matter and that are adapted to fit forklifts and/or other devices/machines to move the pallets. Packages of s products can be stacked on top of the pallets. Warehouses have been designed to permit forklifts to put and pull pallets from racks as needed. fts and other sorts of vehicles move h a warehouse and transport pallets and packages. [0003[ Warehousing, including transportation and storage of s, is labor—intensive, and the labor costs take a large part of operating expenses of a warehouse. The process of reducing labor costs takes careful warehouse labor and occupancy planning, and balancing ef?ciency with maintaining a high level of customer service. Running an inef?cient warehouse may result in bottlenecks in different processes, such as receipt, picking, and g. For example, an inef?ciently planned warehouse may cause workers, forklifts, and other vehicles to walk or run signi?cantly long distances, taking longer time to complete the process.

SUMMARY [0004[ This document generally describes er-based technology for zing a warehouse space, such as warehouse racks. Some embodiments of the technology determines a storage duration of a pallet in a warehouse, and further determines an optimal storage location for the pallet in the warehouse. For example, the technology can determine how long an inbound pallet will stay in a use, and locate an optimal area of the warehouse to store the pallet.

Such an optimal pallet storage area is selected to reduce labor costs in transporting the pallet in, within, and out of the warehouse and further optimize the management of multiple pallets in the warehouse as a whole. In a simple example, the technology can determine that a pallet to be stored in a warehouse is likely to stay for a relatively short period of time (e.g., a couple of days), and then identify a part of warehouse area that is close to an pickup area (e.g., an entrance of the warehouse) from which the pallet is delivered to the use area, thereby reducing a transportation distance of the pallet. On the other hand, if a new pallet is determined to likely stay for a relatively long peiiod of time (e.g, a few ), the technology can determine a part of warehouse area that is rather far from the pickup area, thereby permitting for the other part of warehouse area to remain available for pallets with shorter storage durations so that such pallets with shorter storage ons are carried over shorter ces, y reducing overall labor costs. [0005[ Storage durations of pallets are not typically deterministic due to a variety of dynamic factors, such as suppliers, customers, different storage and/or ry requirements, seasonality of products, regions, timing of storage and delivery, etc. Some embodiments of the technology are con?gured to predict an expected duration during with an d pallet will be stored in a warehouse. Such prediction can be performed using a machine learning algorithm based on a variety of available data, Vitals, and statistics. When an expected storage duration of a pallet is predicted, the technology can determine an optimal area of the use (e.g., a location or section of use racks) to which the pallet will be allocated. Various machine learning algorithms can be used for prediction. Some example machine learning algorithms used for prediction employ arti?cial neural networks, which can include an input layer for ing inputs (e. g., a y of factors described herein), one or more middle layers for processing the inputs, and an output layer for ting pallet storage durations (e. g., percentiles). [0006[ In some implementations, the technology described herein can predict a percentile of an inbound pallet for storage duration in the warehouse. For example, a lower storage percentile of a pallet (e. g., the 10th percentile) may indicate that the pallet stays a shorter period of time (e. g., 10 days) than a pallet with a higher storage percentile (e.g., the 90th percentile, which may indicate 6 month of storage in the use, by way of example). [0007 [ A variety of factors can be used to predict a e duration of a pallet. For example, historical data about storage of pallets of the same or similar kind and/or pallets for the same or similar customer can be used for prediction of a storage duration of a particular pallet. In addition or alternatively, historical data for pallets of different kind and/or pallets for different customers can be used for prediction. In one example, historical performance of pallet storage for the same customer and/or different customers over time can be considered to generate tion of a storage duration of a ular pallet to be stored in the warehouse. Although different customers may have different ements for pallet storage and transportation, the pallet storage and/or transportation for ent customers can show a pattern associated with attributes of the pallets (e. g., types of products in the pallets), and such a pattern can be used to predict a storage duration of a ular pallet (e.g., a storage percentile of the pallet). By way of example, pallets of strawberry products, regardless of whether they are for the same or different customers, may be treated, stored, and distributed similarly to maintain freshness under certain stances. [0008[ Further, the factors used to predict a pallet storage duration can include facility, er, product types, timing (e.g., day of the week, week of the year, etc.) in which pallets come in the warehouse, item description, seasonality, and other le information. At least some of the factors may be identi?ed from a stock keeping unit (SKU). In some implementations, an item description for the product in the pallet (e.g., “6 ounce strawberry puree”) can be parsed to a machine-understandable language using, for example, natural language processing so that the item description can be used to t a storage duration of the pallet in the warehouse. [0009[ Some embodiments of the technology can determine an optimal location of the warehouse that corresponds to the storage duration (e.g., a percentile) of the pallet based on predetermined rules that map pallets storage durations (e.g., percentiles) to areas of the warehouse (e.g, sections of warehouse racks). For example, a ity of storage racks in the warehouse have a plurality of sections corresponding to ranges of pallet e ons. The plurality of sections in each storage rack can be arranged by distance from an area (e. g., a docking area) from which pallets are delivered to the storage rack. By way of example, a storage rack can have multiple sections having a far-front section, a far-end section, and one or more middle sections. The far-front section is closest to the entrance of the warehouse and used to store s having relatively short storage durations (e.g., 0-10 days, or 0—5 percentile of storage duration). The far-end n is farthest to the entrance of the use and used to store pallets having vely long storage durations (e.g., 4-6 months or longer, or 90-100 percentile of storage duration). The middle sections can be arranged between the far-front section and the far-end section, and split to be used for pallets of ent ranges of storage durations between the ones for the far-front section and the ones for the far-end section. [0010[ Some embodiments of the technology can also er a dimension (e.g., a height) of an inbound pallet to determine an optimal storage on for the pallet in a warehouse, such as a warehouse that includes a plurality of storage racks with a plurality of heterogeneous rack openings red to receive pallets with different dimensions (e.g, heights). The optimal storage location for an inbound pallet can be determined in a way to optimize labor cost saving and space utilization, which do not necessarily coincide especially where the warehouse storage racks have ent sizes of rack openings. For example, a storage location of a pallet that matches a storage on of the pallet is not necessarily identical to a storage location that would provide maximum space utilization of the warehouse. In a simple example, a pallet may be determined to be best located in a third column of the warehouse rack from its front based on the pallet’s expected e duration, but the third column may not have a rack opening that fits the size (e.g., height) of the pallet. In this instance, another rack column and/or opening needs to be ined which is optimized with respect to both the storage duration and the pallet size. [0011[ In some implementations, an optimal storage location for a pallet can be determined using a cost function that re?ects matching (or degrees thereof) on pallet storage duration and pallet size. For e, the technology can identify a plurality of candidate rack gs that are available (e. g., the rack openings that are not ed), and calculate an optimization value (e.g., score) for each candidate rack opening based on the storage duration and the size of the . The optimization value can be calculated using a on match value (e.g., score) and a size match value (e.g., score). The duration match value can represent how close the candidate rack opening matches the pallet in terms of the pallet’s storage duration (e.g., percentile). The size match value can represent how close the candidate rack opening matches the pallet in terms of the pallet’s size (e.g., height). [0012[ In some implementations, the duration match value and the size match value can be mentarily weighted, which can be adjusted to meet different needs, such as needs to save labor costs more, or needs to increase the usage (e.g., occupancy) of the warehouse racks. For example, if the size match value is more weighted than the duration match value, the occupancy of the warehouse can increase by ?tting pallets to rack openings with matching sizes (without having to place small pallets to larger rack openings to waste the remaining space therein). On the other hand, if the duration match value is more weighted than the size match value, the labor costs can be saved more by placing pallets to predetermined corresponding sections of the warehouse racks. [0013[ Some embodiments of the technology can determine an occupancy of the warehouse (e.g., use racks), and scale the effective space of the warehouse based on the occupancy.

By way of example, if the warehouse is ed to be occupied only 50%, the ns of the warehouse, in which pallets are ted based on their storage durations and/or sizes, can be selected to be located within only a d portion (e.g., the front half close to the entrance) of the warehouse that is a 50% of the entire space of the warehouse, so that the warehouse is effectively treated as a half size of the warehouse. [0014[ Particular embodiments bed herein include a system for ng a ity of pallets in a warehouse. The system may include a plurality of storage racks having a plurality of rack openings, a database that is programmed to store pallet allocation data that associate expected durations of pallets with a ity of sections of the storage racks, and a computer system including one or more processors that are programmed to perform operations. The ions may include one or more of the following processes: identifying a pallet delivered to the warehouse; determining an expected storage duration of the pallet in the warehouse; determining a storage location of the pallet in the warehouse based on the expected duration of storage; and transmitting ation identifying the storage location to equipment for placement of the pallet. [0015[ In some entations, the system can optionally include one or more of the following features. Determining an expected storage duration of the pallet may include determining a on percentile of the pallet. The expected storage duration of the pallet may be determined based on historical inventory data. The historical ory data may identify pallets stored in the use, times at which the pallets were stored, and durations in which the pallets were stored. Determining an expected storage duration of the pallet may include predicting the expected storage duration using a machine learning based on a plurality of input factors. The input factors may include at least one of a type of items contained in the pallet, a customer of the pallet, a day of week, a week of year, a pallet description, an item description, and historical inventory data. Determining a storage location of the pallet in the warehouse may include determining an area of the warehouse based on a travel distance of the pallet, the travel distance corresponding to the expected on of e. Determining a storage location of the pallet in the warehouse may e determining an area at a distance from an entrance of the warehouse based on the ed duration of e. The plurality of sections may be ed by distance from an entrance of the warehouse. The plurality of sections may be mapped to different pallet duration tiles. The expected storage duration may be identi?ed as a ?rst pallet duration percentile of the pallet duration percentiles. The storage location may be a ?rst section of the plurality of sections, the ?rst section being mapped to the ?rst pallet duration percentile. The plurality of sections may include a ?rst section and a second section being arranged farther from the entrance of the warehouse than the ?rst section, the ?rst section mapped to a ?rst percentile, and the second section mapped to a second percentile greater than the ?rst percentile. The warehouse may include a plurality of storage racks. Each rack may have the plurality of ns, and he plurality of sections may be mapped to different pallet duration percentiles. The plurality of storage racks may include one or more horizontal bars adjustable along a ity of elevations on the storage racks to de?ne a plurality of rack openings within the storage racks. The operations may include determining a height of the . Determining a e location of the pallet in the warehouse may include determining one of the ity of sections in the storage racks and one of the rack openings in the storage racks based on a cost function of the expected duration of storage and the height of the pallet.

The ions may include determining a height of the pallet, identifying a plurality of candidate rack openings that are available in the storage racks, calculating optimization values for the candidate rack openings based on the expected duration of storage and the height of the pallet, and determining a rack opening from the ate rack openings having an optimization value exceeding a threshold value, the rack opening being the storage location for the pallet.

Each of the zation values may include a combination of a duration match value and a height match value for each of the candidate rack openings. The duration match value for a candidate rack opening may represent proximity between a section of the storage racks suited for the expected storage duration of the pallet and a section of the storage racks to which the candidate rack opening belongs. The height match value for the candidate rack opening may represent ity between the height of the pallet and a height of the candidate rack opening.

The duration match value and the height match value may be mentanly weighted. The equipment may include a forklift that includes a user interface mmed to automatically output the information identifying the storage location in response to receiving the transmitted information. [0016[ Particular embodiments described herein include a method for managing a plurality of pallets in a warehouse. The method may include identifying a pallet delivered to the warehouse, determining an expected storage duration of the pallet in the use, determining a e location of the pallet in the warehouse based on the expected duration of storage, and transmitting information identifying the storage location to ent for placement of the pallet. [0017 [ In some entations, the method can optionally e one or more of the following features. The warehouse may include a plurality of sections arranged by distance from an entrance of the warehouse. The plurality of sections may be mapped to different pallet duration percentiles. The expected storage duration may be identi?ed as a ?rst pallet on percentile of the pallet duration percentiles. The e on may be a ?rst section of the plurality of sections. The ?rst section may be mapped to the ?rst pallet duration percentile. The plurality of sections may include a ?rst section and a second section being arranged farther from the entrance of the warehouse than the ?rst section. The ?rst section may be mapped to a ?rst percentile, and the second section may be mapped to a second percentile greater than the ?rst percentile. The warehouse may include a plurality of storage racks. Each rack may have the plurality of sections. The plurality of sections may be mapped to different pallet duration percentiles. The plurality of storage racks may include one or more horizontal bars able along a plurality of elevations on the storage racks to de?ne a plurality of rack openings within the storage racks. The method may further include determining a height of the pallet.

Determining a storage on of the pallet in the warehouse may e determining one of the plurality of sections in the storage racks and one of the rack openings in the storage racks based on a cost function of the expected duration of storage and the height of the pallet. The method may further include determining a height of the pallet, identifying a plurality of candidate rack openings that are ble in the e racks, calculating optimization values for the candidate rack openings based on the expected duration of storage and the height of the pallet, and ining a rack opening from the candidate rack openings having an zation value exceeding a threshold value, the rack opening being the storage location for the pallet. Each of the optimization values may include a combination of a on match value and a height match value for each of the candidate rack openings. The duration match value for a ate rack opening may represent proximity between a n of the storage racks suited for the expected storage duration of the pallet and a section of the storage racks to which the candidate rack opening belongs. The height match value for the candidate rack opening may represent proximity between the height of the pallet and a height of the candidate rack opening. The duration match value and the height match value may be complementarily weighted. [0018[ The s, system, and techniques described herein may provide several advantages. For example, some embodiments of the technology provides a solution to ine or t storage durations of pallets (e.g., pallet velocities) based on a variety of factors, and select areas of use racks or other pallet storage location which can optimize pallet travel times and distances in transporting the pallets in and out of the areas and thus save labor costs in managing the facility. Further, some embodiments of the technology can further provide a solution to increase space utilization of the warehouse racks or other pallet storage location by taking into account the sizes of the pallets and the sizes of the rack openings available for the pallets. The technology can determine rack openings that can optimally ?t the pallets in terms of their sizes and further meet the storage durations of the pallets in a way to balance often- atible needs for saving labor costs and maximizing facility space usage. [0019[ The details of one or more implementations are set forth in the accompanying drawings and the description below. Other features and advantages will be apparent from the description and drawings, and from the claims.

BRIEF DESCRIPTION OF THE GS [0020[ illustrates an example system for optimizing pallet storage locations in a warehouse. [0021[ is a ?owchart of an example process for determining an optimal storage location of a pallet in a warehouse and placing it in the determined location. [0022[ illustrates an example chart that shows a distribution of pallets with different storage durations. [0023[ illustrates an example process for determining an optimal storage location of a new pallet based on an expected storage duration of the new pallet. [0024[ illustrates an example system for optimizing pallet storage locations in a warehouse. [0025[ is a ?owchart of an example process for determining an optimal storage location of a pallet in a warehouse and placing it in the determined location. [0026[ is a ?owchart of an e process for selecting an optimal rack opening for a pallet in a warehouse. [0027 [ illustrates an example process for determining an l e location of a new pallet based on an expected storage duration of the new pallet. [0028[ illustrates an example technology for scaling a pallet e area depending on an ncy of the area. [0029[ is a block diagram of computing devices that may be used to implement the systems and methods described in this document, as either a client or as a server or plurality of SCTVCIS.

DETAILED DESCRIPTION OF RATIVE EMBODIMENTS [0030[ In general, this document bes computer-based techniques for optimizing storage ons at a storage facility (e.g., a warehouse) and directing the putaway of inbound s to optimal locations at the storage facility, such as optimal locations in storage racks. [0031[ illustrates an example system 100 for optimizing pallet e ons in a warehouse 102. The warehouse 102 can include a pallet storage area 104 for storing pallets.

The pallet storage area 104 can include a plurality of warehouse racks 106 having rack openings 108 con?gured to receive pallets for e. In this e, at least some of the use racks 106 are homogeneous warehouse racks which have the rack openings of the same size.

The warehouse 102 can further include a loading area 110 from which pallets 114 are delivered to the pallet storage area 104 or the warehouse racks 106. The loading area 110 may be a g dock or bay (or an area close to the loading dock or bay) where pallets are unloaded from, or loaded onto, vehicles 112, such as trucks, trains, or other suitable vehicles. The loading area 110 may be other areas in or outside the warehouse, which pallets are temporarily placed before delivered to the pallet storage area 104 or the warehouse racks 106. The pallets 114 can be transported from the loading area 110 using various methods. In one example, forklifts 116 can be operated to pick up a pallet 114 and transport the pallet to a desired location in the pallet e area 104 or a desired rack opening 108 of a storage rack 106. [0032[ The example system 100 r includes a computer system 120 that is programmed to determine l locations for pallet storage in the pallet storage area 104 (e.g., rack openings in the warehouse racks 106) based at least in part on pallet data 122 and/or historical inventory data 124. The pallet data 122 include information of s to be stored in the pallet storage area 104, such as type of items in pallets, item description, pallet description, customers, ties, timing (e.g., day of the week, week of the year, etc.) in which pallets come in the warehouse, seasonality, etc. [0033[ The historical inventory data 124 include historical data about pallets (e.g., pallets of the same or similar kind and/or pallets of ent kinds) which have been delivered in and out, and stored in, the pallet storage area 104 over time. The historical inventory data 124 can include, for example, histon'cal performance of pallet storage for the same customer and/or different customers over time can be considered to generate prediction of a storage on of a particular pallet to be stored in the warehouse. Different customers may have different requirements for pallet storage and transportation. However, the pallet e and/or transportation for different ers can show a n associated with attributes of the pallets (e.g., types of products in the pallets). By way of example, pallets of erry products, regardless of whether they are for the same or different customers, may be treated, stored, and distributed similarly to maintain ess under certain circumstances. [0034[ The computer system 120 can determine an expected duration in which a pallet, such as a pallet 114 at the loading area 110, is stored in the pallet storage area 104 (e.g., a warehouse rack 106). The computer system 120 can use the pallet data 122 and/or the historical inventory data 124 to determine an expected pallet storage duration of a particular pallet, which can ent how long the pallet will stay for storage in the pallet storage area 104, or how soon the pallet will be removed from the pallet storage area 104 after being placed in the pallet storage area 104 (i.e., a pallet ty). [0035[ The computer system 120 can use pallet allocation data 126 to determine optimal locations to store pallets in the pallet storage area 104. The pallet allocation data 126 can include rules for ting identi?ed pallets to desired locations (e.g., rack openings) in the pallet storage area 104. As described herein, the pallet allocation data 126 can include information that maps expected e durations to the pallet storage area 104. [0036[ For example, the pallet allocation data 126 include information (e.g., a data table) that associates different pallet storage durations with different sections of the use racks 106 in a way to reduce labor costs for transporting pallets to and from the warehouse racks 106. As illustrated in the warehouse racks 106 can have different sections 140 (such as 140A-D) that are divided and arranged at ent distances L (such L1-L4) from the source of pallets (e.g., the g area 110). For example, ?rst, second, third, and fourth sections 140A-D of the warehouse racks 106 are arranged at ?rst, second, third, and fourth distances L1—L4, respectively, from the loading area 110 with the ?rst section 140A being arranged closest to the loading area 110 and the fourth section 140D being ed farthest from the loading area 110.

The pallet allocation data 126 provides assignment of the ?rst, second, third, and fourth sections 140A-D to different ranges of pallet storage durations, such as 0-10 day storage, 10—30 day storage, 30-90 day storage, and 90-200 day storage, respectively. In this e, therefore, based on the pallet allocation data 126, an inbound pallet that is determined to have a storage duration of 3 days will be assigned to the ?rst section 140A. Similarly, this example of the pallet allocation data 126 assigns a pallet having a storage on of 120 days to the fourth section 140D. While the warehouse racks 106 are illustrated to have four sections 140 herein, the warehouse racks 106 can have less than or more than four sections. In some implementations, the warehouse racks 106 can have 100 sections or more so that each integer percentile of pallet storage duration has its own category and assignment to one or more sections of the warehouse racks. By way of example, 5th tile pallets can be stored in any of one or more sections corresponding to such pallets in a way that the pallets can turn more quickly than higher percentile (e.g., 50th percentile) pallets. In this way, the same percentile pallets do not always need to be stored in the same on of rack. [0037 [ The computer system 120 can include a server system, cloud-based computer system, desktop, , mobile ing device, other computing device/system. In some entations, the er system 120 can be con?gured as a Warehouse Management System (also referred to as “WMS”) that is a specialized computer system to manage storage and retrieval of inventory at the facility, and to interface with devices and components within the facility, such as forklifts, sensors, HVAC systems, lighting systems, and/or other devices and ents. For example, the WMS can determine where pallets arriving at the ty should be placed (e.g., identifying rack opening for pallets), manage and track the locations of pallets, identify pallets that should be removed for shipment out of the ty, and communicate with forklifts (and other devices) to provide pick and place information for pallets. [0038[ Referring still to the system 100 can identify the characteristics of pallets that are to be stored in a warehouse, determine optimal storage locations of the pallets in the warehouse, and transport the s to the determined storage locations for storage. For example, the computer system 120 can receive information about a new pallet 114 (Step A). In some implementations, the information about the new pallet 114 can be transmitted from a forklift 116 (or other vehicle) that picks up the new pallet for transportation. The new pallet 114 or the items thereon (e.g., a SKU) can be scanned before or after it is picked up by the forklift 116, and the scanned information can be transmitted to the computer system 120. In other implementations, the new pallet information can be obtained in other ways, such as provided by a supplier of the new pallet when the new pallet is delivered to the warehouse. [0039[ The computer system 120 can identify the new pallet 114 based on the received new pallet information (Step B). The computer system 120 can obtain pallet information about the new pallet 114 (Step C). For example, the computer system 120 can retrieve the pallet information from the pallet data 122. The pallet information includes facility, customer, product types, timing (e.g., day of the week, week of the year, etc.) in which pallets come in the warehouse, item description, seasonality, and other suitable information. [0040[ The computer system 120 can determine an ed e duration of the new pallet 114 (Step D). The expected e duration of the new pallet 114 can be determined based at least in part on the pallet information (e.g, the pallet data 122) and/or the historical inventory data ated with the warehouse. The computer system 120 can determine an optimal storage on according to the expected storage duration of the new pallet 114 (Step E). The optimal storage location can be ined using the pallet allocation data 126 that can identify a section of the warehouse rack that corresponds to the expected storage duration of the new pallet 114. The computer system 120 can transmit the optimal storage location to the ft 116 (Step F), and the ft 116 can transport the new pallet 114 and place it to the optimal storage location (e.g, the determined section of the warehouse rack) (Step G). [0041[ is a ?owchart of an example process 200 for determining an optimal storage location of a pallet in a warehouse, and placing it in the determined location. The process 200 can begin by receiving new pallet data (Block 202). The new pallet data can be used to identify a new pallet that needs to be stored in a warehouse. The new pallet data can be ed by a supplier of the new pallet as the pallet is orted to the warehouse. Alternatively, the new pallet data can be obtained ly (e.g., by a worker who identi?es the new pallet and enters information about the pallet (e. g., a SKU) via a terminal that is communicatively connected to one or more computing devices, such as the computer system 120. Alternatively, the new pallet data can be obtained by scanning the pallet (e.g, scanning the barcode on the pallet) before or while the new pallet is moved by a forklift or other suitable vehicles in the warehouse. The new pallet data can be saved as the pallet data 122 in [0042[ The s 200 can further include obtaining a variety of factors 230 associated with the new pallet (Block 204). The factors include one or more of historical inventory information, pallet item attributes (e. g., product types, seasonality, etc.), customer of the pallet, timing ation (e. g., day of the week, week of the year, etc. in which the pallet is delivered in and out the warehouse), facility ation (e.g., type of facility, operational conditions, etc.), pallet description, pallet item description, and other relevant information. [0043[ Historical inventory information can include ation storage of pallets of the same or similar kind and/or pallets for the same or similar customer over a predetermined period of time in the past, or over the entire time of storage at the facility in the past. In addition or alternatively, the historical inventory information can include historical data for pallets of different kind and/or pallets for different customers can be used for prediction. Historical WO 02111 2020/061235 performance of pallet storage for the same customer and/or different customers over time can be considered to generate tion of a storage duration of a particular pallet to be stored in the warehouse. [0044[ Item description can indicate a human-readable description about items in the pallet, such as “6 ounce strawberry puree.” The item description can be parsed to a machine- understandable language using, for example, natural ge processing so that the item ption can be used to predict a storage duration of the pallet in the warehouse. [0045[ The process 200 can include predicting an expected e duration of the new pallet (Block 206). For example, the expected storage duration can be predicted using a machine learning algorithm. Various machine learning algorithms can be used for prediction. Some example machine learning algorithms used for prediction employ arti?cial neural networks with multiple layers, such as an input layer, one or more middle layers, and an output layer. In some implementations, prediction of the expected storage duration can be performed by the input layer that es a variety of factors associated with the new pallet as inputs (Block 220), the middle layers that process the inputs (Block 222), and the output layer that generates a storage duration of the pallet (Block 224). As described herein, the storage duration of the pallet can be represented as a percentile of the pallet for storage duration in the warehouse. In some implementations, the process 200 can t a distribution for the set of input data (e.g., input factors), randomly sample from the distribution, and then convert to a percentile after ranking that on relative to durations of other pallets in the room. [0046[ The process 200 can include determining an optimal storage location that corresponds to the predicted storage duration based on pallet allocation data 240 (Block 208). The pallet allocation data 240 can be saved as the pallet allocation data 126 in In some implementations, the pallet allocation data 240 provide rules as a table that maps ranges of pallet storage durations (e.g., percentiles) with storage ons (e.g., sections of a rack). By way of example, if a new pallet is predicted to have 25 tile of storage duration, the new pallet is determined to be placed in Section 2 of the rack, using the illustrated pallet allocation data 240. [0047 [ While the pallet allocation data 240 is rated herein to have four sections 140 for four different percentile ranges, the warehouse racks 106 can be divided up into a large number of smaller ns (e.g, 100 or more smaller sections for each rack) so that the sections of the racks are arranged to be continuous per corresponding percentiles of pellets, and allocations of pallets to the sections are ?exible. In some entations, a warehouse rack can have 100 sections or more so that each integer percentile of pallet storage duration has its own category and assignment to one or more sections of the rack. By way of example, 5th percentile s can be stored in any of one or more ns ponding to such pallets in a way that the pallets can turn more quickly than higher percentile (e. g., 50th percentile) pallets. In this way, the same percentile pallets do not always need to be stored in the same exact location of rack. [0048[ The process 200 can include transmitting the determined storage location to transportation equipment, such as a forklift, so that the new pallet is transported and placed to the determined storage location for storage in the warehouse (Block 210). [0049[ illustrates an example chart 300 that shows a bution of pallets with different storage durations. The chart 300 shows a number of pallets 302 (in a vertical axis) for different storage durations 304 (in a horizontal axis). The chart 300 further shows example percentiles 310 and multiple percentile ranges 320. For example, the chart 300 depicts example percentiles (e.g., a ?rst percentile 310A, a second percentile 310B, and a third percentile 310C), and example percentile ranges (e.g., a ?rst range 320A, a second range 320B, a third range 320C, and a fourth range 320D). By way of example, if a new pallet is predicted to have a storage duration of 23 days, the storage duration of the new pallet is between the ?rst percentile 310A and the second percentile 310B, thereby falling within the second range 320B. [0050[ The extent of e ons for determining percentiles can be determined at various points of time. In one e, the percentile of a new pallet can be ined based on the entire pallets that are stored in the storage area (e.g., the use racks) at the time that the new pallet is identi?ed to be stored in the storage area. Alternatively, the percentile of a new pallet can be determined based on the pallets that have been historically stored in the storage area over either the entire time of operating the warehouse or a predetermined period of time.

Alternatively, the percentile of a new pallet can be ined based on an expected future inventory of pallets, such as at a predetermined future time after the new pallet is identi?ed to be stored in the storage area. [0051[ illustrates an e process 400 for determining an optimal storage location of a new pallet based on an expected storage duration of the new pallet. In this example, a storage rack 402, as presented in its side view, is arranged such that each rack g 404 has the same size, as indicated by the consistent shelf height across the length of the rack. The storage rack 402 has a ity of sections 406 (including 406A-D) arranged at different distances from a loading area 408 where s 410 are ready to be transported to the storage rack 402. In the illustrated example, the e rack 402 has four sections arranged by distance from the loading area 408. In particular, ?rst, second, third, and fourth sections 406A-D are ed at ?rst, second, third, and fourth distances L1-L4, respectively, from the loading area 408. The ?rst distance L1 is shorter than the second distance L2, the second distance L2 is shorter than the third distance L3, and the third distance L3 is shorter than the fourth distance L4. [0052[ Each of the rack sections 406 is assigned to a particular range of pallet storage duration. Such assignment can be predetermined and stored as pallet allocation data or rules 412 (e. g., the pallet allocation data 240). In some implementations, a short storage duration of a pallet indicates a short turnaround of the pallet (e. g., the pallet is delivered in and out of the warehouse within a short period of time). Thus, a shorter storage duration can be mapped to a rack section 406 that is closer to the loading area 408 so that racks with quicker turnaround can be stored closer to the loading area 408, thereby saving labor costs in ring in and out of the warehouse rack. By way of example, as illustrated, the ?rst n 406A is assigned to a pallet storage duration between 0 and 5 percentile, the second n 406B is assigned to a pallet storage duration between 5 (including 5) and 50 percentile, the third section 406C is assigned to a pallet storage duration between 50 (including 50) and 90 percentile, and the fourth section 406D is assigned to a pallet storage on between 90 (including 90) and 100 percentile. [0053[ The pallets 410 that are delivered from the loading area 408 can be identi?ed, and an expected storage on of each pallet is determined as described herein. In the illustrated example, a ?rst pallet 410A is determined to have a storage duration of 2 percentile, and thus is delivered and stored in the ?rst n 406A of the rack 402. A second pallet 410B is ined to have a storage on of 25 percentile, and thus is stored in the second section 406B of the rack 402. A third pallet 410C is determined to have a storage duration of 60 percentile, and thus is stored in the third section 406C of the rack 402. A fourth pallet 410D is determined to have a storage duration of 95 percentile, and thus is stored in the fourth section 406D of the rack 402. [0054[ illustrates an example system 500 for optimizing pallet storage locations in a use 102. The system 500 is con?gured similarly to the system 100 in For example, the system 500 can include a warehouse 502, a pallet storage area 504, a plurality of warehouse racks 506 with rack openings 508, a loading area 510, vehicles 512, and a computer system 520, which are similar to the warehouse 102, the pallet storage area 104, the plurality of warehouse racks 106 with the rack openings 108, the loading area 110, the es 112, the computer system 120 in the system 100. [0055[ In this example, at least some of the warehouse racks 506 are heterogeneous warehouse racks having rack openings 508 of different sizes. The rack gs 508 can be varied to increase or maximize the use of the rack space (or to reduce or ze the unused space) for the speci?c storage ty. In some implementations, the rack openings can be adjusted over time as the use of the facility changes over time (e.g., different products are stored at the warehouse, different companies or clients are using the warehouse). [0056[ The computer system 520 is con?gured to determine optimal locations for pallet storage in the pallet storage area 504 (e.g., rack openings in the warehouse racks 506) based at least in part on pallet data 522, historical inventory data 524, and/or pallet size data 528.

Similarly to the pallet data 122 in the pallet data 522 include information of pallets to be stored in the pallet storage area 504. Similarly to the historical inventory data 124 in the historical ory data 524 include historical data about pallets which have been delivered in and out, and stored in, the pallet storage area 504 over time. The pallet size data 528 include information of sizes of s to be stored in the pallet storage area 504, such as heights, widths, lengths, volumes, etc. In some implementations, the pallet size data 528 can be included in the pallet data 522. Similarly to the pallet data, the pallet size data 528 can be provided by a supplier of the new pallet as the pallet is orted to the warehouse. Alternatively, the pallet size data 528 can be obtained manually (e.g., by a worker who identi?es the pallet and enters information about the pallet (e.g., a SKU) via a terminal that is communicatively connected to one or more computing devices. Alternatively, the pallet size data 528 can be obtained by scanning the pallet (e.g, scanning the barcode on the pallet) before or while the new pallet is moved by a forklift or other suitable es in the warehouse. [0057 [ The computer system 520 can determine an expected duration in which a pallet, such as a pallet 514 at the loading area 510, is stored in the pallet storage area 504 (e.g., a warehouse rack 506). The computer system 520 can use the pallet data 522 and/or the historical inventory data 524 to determine an expected pallet storage duration of a ular pallet, which can represent how long the pallet will stay for storage in the pallet storage area 504, or how soon the pallet will be removed from the pallet storage area 504 after being placed in the pallet storage area 504 (i.e., a pallet velocity). [0058[ Further, the er system 520 can determine a size of the pallet to be stored in the pallet storage area 504 (e.g., a use rack 506). The computer system 520 can retrieve the size of the pallet from the pallet size data 528. Alternatively, the computer system 520 can use the pallet data 522, the historical inventory data 524, and/or other suitable data to determine or predict the size (e.g., height) of a particular pallet. [0059[ The computer system 520 can determine optimal locations to store pallets in the pallet storage area 504 based in part on pallet allocation data 526, the pallet size data 528, and/or rack opening data 530. The pallet allocation data 526 can include rules for allocating identi?ed pallets to desired locations (e.g., rack openings) in the pallet storage area 504. As described herein, the pallet allocation data 526 can include information that ates expected storage durations and/or sizes of pallets to locations (e.g, sections or rack gs) in the pallet storage area 504. For example, similarly to the pallet allocation data 126, the pallet allocation data 526 include information (e.g., a data table) that associates different pallet storage durations with different sections of the warehouse racks 506 in a way to reduce labor costs for transporting pallets to and from the warehouse racks 506. As rated in the warehouse racks 506 can have different sections 540 (such as ) that are d and arranged at different distances L (such L1 l-Ll4) from the source of pallets (e.g., the loading area 510). For example, ?rst, second, third, and fourth sections 540A-D of the warehouse racks 506 are arranged at ?rst, second, third, and fourth distances Ll l-Ll4, respectively, from the loading area 510 with the ?rst n 540A being ed closest to the loading area 510 and the fourth section 540D being arranged farthest from the loading area 510. [0060[ In on to the pallet allocation data 526, the computer system 520 can further use the pallet size data 528 and the rack g data 530 to determine optimal storage locations for s in the storage racks 506 with rack openings 508 of ent sizes. The rack opening data 530 can include information about each rack opening, such as locations (e.g., distances from a particular area, such as a loading area), sections (e.g, percentile sections), sizes (e.g., heights), bility (e. g., whether the rack opening is currently being used or unused), and other data relating to pallet storage. [0061[ Considering expected storage duration and sizes of pallets, the computer system 520 can determine storage ons of the pallets that optimize both of labor cost saving and space utilization. In some implementations, the computer system 520 can determine an optimal storage location of a pallet using a cost function that re?ects both pallet storage duration matching and pallet size matching. For example, the computer system 520 can ate a cost value or score of each potential storage location (e.g., a rack opening) that is a function of pallet storage duration ng and pallet size matching. The pallet storage duration matching can indicate how close a potential e location matches a particular pallet in terms of the pallet’s storage duration (e.g., percentile). The pallet size matching can indicate how close a potential storage location s the pallet in terms of the pallet’s size (e.g., height). In some entations, the components of the cost function can be weighted. For example, values of the pallet storage on matching and the pallet size matching can be selectively weighted to re?ect different needs, such as requiring more labor cost saving or more space utilization. For example, if the pallet size ng is more weighted than the pallet storage duration matching, the occupancy of the warehouse can increase by ?tting pallets to rack openings with matching sizes (without having to place small pallets to larger rack openings to waste a space therein). On the other hand, if the pallet storage duration matching is more weighted than the pallet size matching, the labor costs can be saved more by placing s to predetermined corresponding ns of the warehouse racks. [0062[ Referring still to the system 500 can identify the teristics of pallets that are to be stored in a warehouse, determine l storage locations of the pallets in the warehouse, and transport the pallets to the determined storage locations for storage. For example, the computer system 520 can receive information about a new pallet 514 (Step A). In some implementations, the information about the new pallet 514 can be transmitted from a forklift 516 (or other vehicle) that picks up the new pallet for transportation. The new pallet 514 or the items thereon (e.g., a SKU) can be scanned before or after it is picked up by the forklift 516, and the scanned information can be itted to the computer system 520. In other implementations, the new pallet information can be obtained in other ways, such as provided by a supplier of the new pallet when the new pallet is delivered to the use. [0063[ The computer system 520 can fy the new pallet 514 based on the received new pallet information (Step B). The computer system 520 can obtain pallet information about the new pallet 514 (Step C). For example, the computer system 520 can retrieve the pallet information from the pallet data 522. The pallet information includes facility, customer, product types, timing (e.g., day of the week, week of the year, etc.) in which pallets come in the warehouse, item description, seasonality, and other suitable information. [0064[ The computer system 520 can determine an expected storage duration of the new pallet 514 (Step D). The expected storage duration of the new pallet 514 can be determined based at least in part on the pallet information (e.g., the pallet data 522) and/or the historical inventory data associated with the use. The computer system 520 can determine a size (e.g., height) of the pallet 514 (Step E). For e, the computer system 520 can obtain the pallet size from the pallet size data 528, which may or may not be part of the pallet data 522. [0065[ The computer system 520 can determine an optimal storage location according to the expected storage duration and the size of the new pallet 514 (Step F). The optimal storage on can be determined to optimize the labor costs (e.g., based on the ed storage duration) and the space usage (e.g., based on the pallet size) of the storage racks. The computer system 520 can transmit the optimal storage location to the forklift 516 (Step G), and the forklift 516 can transport the new pallet 514 and place it to the l storage location (e.g., the determined section of the warehouse rack) (Step H). [0066[ is a ?owchart of an e process 600 for determining an optimal storage location of a pallet in a warehouse, and placing it in the determined location. The process 600 can begin by receiving new pallet data (Block 602). The new pallet data can be used to identify a new pallet that needs to be stored in a warehouse. The new pallet data can be provided by a supplier of the new pallet as the pallet is transported to the warehouse. Alternatively, the new pallet data can be obtained manually (e.g., by a worker who identi?es the new pallet and enters information about the pallet (e.g., a SKU) via a terminal that is communicatively connected to one or more computing s, such as the computer system 120. Alternatively, the new pallet data can be obtained by scanning the pallet (e.g., scanning the barcode on the pallet) before or while the new pallet is moved by a forklift or other suitable vehicles in the warehouse. The new pallet data can be saved as the pallet data 522 in [0067 [ The process 600 can further include obtaining a variety of factors 630 associated with the new pallet (Block 604). The s include one or more of historical inventory information, pallet item attributes (e. g., product types, seasonality, etc), customer of the pallet, timing information (e.g, day of the week, week of the year, etc. in which the pallet is delivered in and out the warehouse), facility information (e.g., type of facility, operational conditions, etc), pallet ption, pallet item description, pallet size, and other relevant information. [0068[ Historical ory information can include information storage of pallets of the same or similar kind and/or s for the same or similar customer over a predetermined period of time in the past, or over the entire time of storage at the facility in the past. In addition or atively, the historical inventory information can include historical data for pallets of different kind and/or pallets for different customers can be used for prediction. Historical performance of pallet storage for the same customer and/or different customers over time can be considered to generate prediction of a e duration of a ular pallet to be stored in the warehouse. [0069[ Item desc?ption can indicate a human-readable description about items in the pallet, such as “6 ounce strawberry puree.” The item description can be parsed to a machine- understandable ge using, for example, natural language processing so that the item description can be used to predict a e duration of the pallet in the warehouse. [0070[ The process 600 can include predicting an expected storage duration of the new pallet (Block 606). For example, the expected storage duration can be predicted using a machine learning thm. Various machine learning algorithms can be used for tion. Some example machine learning algorithms used for tion employ arti?cial neural networks with multiple layers, such as an input layer, one or more middle layers, and an output layer. In some implementations, prediction of the expected storage duration can be performed by the input layer that receives a variety of factors associated with the new pallet as inputs (Block 620), the middle layers that process the inputs (Block 622), and the output layer that generates a storage duration of the pallet (Block 624). As described herein, the storage duration of the pallet can be represented as a percentile of the pallet for storage duration in the warehouse. [0071[ The process 600 can include determining a size of the new pallet (Block 608). A height of the new pallet can be primarily considered to determine whether and/or how the new pallet can fit in a rack opening of a storage rack. In some implementations, the size of the new pallet can be obtained from the factors 630. [0072[ The s 600 can include determining an optimal storage location using a cost function (Block 610). Some example cost function is configured to optimize a labor cost for ring pallets to and from a pallet e area (e.g., an area where one or more racks are d), and a cost for utilizing the space of the pallet e area (e.g., a cost for using the racks with used and unused rack openings). In some implementations, with respect to a particular pallet, the cost for each location in the pallet storage area, such as each rack opening (or each unoccupied rack opening) in the storage racks, can be calculated based on storage duration matching and storage size matching. Storage on matching relates to how suitable the location (e.g., rack opening) in the pallet storage area is for an expected storage duration (e.g., percentile) of the pallet to save labor costs. Storage size matching relates to how suitable the location (e.g., rack opening) in the pallet storage area is for the size (e.g., height) of the pallet to maximize the space utilization in the storage area. [0073[ Storage duration matching can be determined using pallet allocation data 640, The pallet allocation data 640 can be saved as the pallet allocation data 526 in In some implementations, the pallet allocation data 640 e rules as a table that maps ranges of pallet storage ons (e.g., percentiles) with storage locations (e.g., ns of a rack). By way of example, if a new pallet is predicted to have 25 percentile of storage duration, the new pallet is determined to match Section 2 of the rack, using the illustrated pallet allocation data 640. [0074[ e size matching can be determined using rack opening data 642. The rack opening data 642 can saved as the rack g data 530 in The rack opening data 642 can provide the size (e.g., height) of each rack opening in one or more racks. The rack opening data 642 can further provide availability of each rack opening, such as r the rack opening is open to store a pallet, or has been occupied and not available for another pallet. [0075[ The process 600 can include transmitting the determined storage on to transportation equipment, such as a forklift, so that the new pallet is transported and placed to the determined storage location for storage in the warehouse (Block 612). [0076[ is a ?owchart of an example process 700 for selecting an optimal rack opening for a pallet in a warehouse. The s 700 can include identifying a plurality of candidate rack openings in one or more storage racks (Block 702). The ate rack openings can be rack openings that are not currently occupied and thus ble to store pallets. The candidate rack openings can be determined based on rack opening data, such as the rack opening data 530, 642. [0077 [ The process 700 can include calculating optimization values for the candidate rack openings (Block 704). In some implementations, an optimization value for each of the candidate rack openings can be calculated by determining a storage on match value (Block 710), determining a storage height match value (Block 712), and calculating the optimization value based on the storage duration match value and the storage height match value (Block 716). The storage duration match value can indicate whether an expected storage duration of the pallet matches a storage duration range that is assigned to the candidate rack opening, or how close the expected storage duration of the pallet is to the storage on range assigned to the candidate rack opening. The e height match value can indicate whether the height of the pallet matches a height of the candidate rack opening, or how close the pallet ?ts in the candidate rack opening (e.g., how close the height of the pallet is to the height of the candidate rack opening). [0078[ The optimization value can be calculated using a predetermined function of the storage duration match value and the e height match value. For e, the optimization value can be a sum of the e duration match value and the storage height match value. [0079[ The storage duration match value, the storage height match value, and the zation value can be represented in various forms and/or . For example, such values can be represented as scores either scaled or unscaled. In other es, the optimization values can be represented as binary numbers, indicators, symbols, etc. [0080[ In some implementations, the storage duration match value and/or the storage height match value can be weighted to calculate the optimization value (Block 714). Weights on the storage duration match value and/or the storage height match value can be adjusted to meet various purposes. For example, if labor cost saving is more of concern, the storage duration match value may be more weighted than the storage height match value. If space utilization if more of concern, the storage height match value may be more weighted than the storage duration match value. An example weighted cost function for an optimization value can be: f (x) = CXOC) + (1 - C)Y(X) where x is a pallet, fis an zation value, X is a storage duration match value, Y is a height match value, and C is a weight value. [0081[ The process 700 can include selecting a rack opening for the pallet from the candidate rack gs based on the calculated optimization values (Block 706). For example, one of the candidate rack openings that has the highest optimization value can be considered as an optimal rack opening and selected as the rack opening for the pallet. In other examples, multiple ones of the ate rack openings that have zation values that meet a threshold value (e.g., greater than the threshold value) can be considered as multiple optimal rack openings, and any one of the multiple optimal rack openings can be selected as the rack opening for the pallet.

Depending on the cost on and/or scoring scheme, the optimal rack openings can have either highest or lowest optimization values, or values that are either greater or less than the threshold value. [0082[ In addition to, or alternatively to, the duration match and/or the height match, other factors can also be used to calculate optimization values and/or selecting rack openings. For example, the optimization values can be calculated, and/or rack openings can be selected, such that identical items are placed together or close each other in the rack. In another example, the optimization values can be ated, and/or rack gs can be selected, such that items that are to be d together are placed to be grouped in the rack. Other factors are also possible. [0083[ In some entations, p-norm can be used for calculation. The ion here is “calculating distance over multiple metrics.” Various types of norm can be used according to the value of p. For example, p=1 is “Manhattan/Taxicab” distance, p=2 is “as the crow ?ies” distance, and p=3 is “as the crow ?ies but avoiding all McDonalds franchises.” Other p-values are used for calculation in other examples. In other implementations, other ce metrics can be used for calculation. [0084[ illustrates an example process 800 for determining an optimal storage location of a new pallet based on an expected storage duration of the new pallet. In this example, a storage rack 802, as presented in its side view, is arranged such that the rack openings 804 are varied to increase or maximize the use of the rack space, or to reduce or minimize the unused space, for the speci?c storage facility. In some implementations, the rack openings 804 can be adjusted over time as the use of the facility changes over time (e.g., different products are stored at the warehouse, different companies or clients are using the warehouse). [0085[ The storage rack 802 has a plurality of sections 806 (including 806A—D) arranged at different distances from a loading area 808 where pallets 810 are ready to be transported to the storage rack 802. In the illustrated example, the storage rack 802 has four sections arranged by distance from the loading area 808. In particular, ?rst, second, third, and fourth ns 806A-D are arranged at ?rst, second, third, and fourth distances 4, respectively, from the loading area 808. The ?rst ce L11 is shorter than the second distance L12, the second distance L12 is r than the third distance L13, and the third distance L13 is shorter than the fourth distance L14. [0086[ Each of the rack sections 806 is assigned to a particular range of pallet storage duration. Such assignment can be predetermined and stored as pallet allocation data or rules 812 (e.g., the pallet allocation data 640). In some entations, a short storage duration of a pallet indicates a short turnaround of the pallet (e.g, the pallet is red in and out of the warehouse within a short period of time). Thus, a shorter storage duration can be mapped to a rack n 806 that is closer to the loading area 808 so that racks with quicker ound can be stored closer to the loading area 808, thereby saving labor costs in delivering in and out of the use rack. By way of example, as illustrated, the ?rst section 806A is assigned to a pallet storage duration between 0 and 5 percentile, the second section 806B is assigned to a pallet e duration between 5 (including 5) and 50 percentile, the third n 806C is assigned to a pallet storage duration between 50 (including 50) and 90 percentile, and the fourth section 806D is assigned to a pallet storage duration between 90 (including 90) and 100 percentile. [0087 [ The pallets 810 that are to be delivered to the rack can be identi?ed, and allocation data 820 of each pallet can be determined as described herein. For example, the allocation data 820 of a ular pallet can include an ed storage duration of the pallet (e.g., in terms of percentile) and a size (e.g., height) of the pallet. [0088[ When a particular pallet to be red and stored in the rack is identi?ed, rack opening data 830 for each rack g 804 that is unoccupied and thus potentially available for the pallet is generated as described herein. For e, the rack g data 830 can include an optimization value of each rack opening with t to the pallet. For example, the optimization value can be calculated based on a storage duration match value and a storage height match value, which may or may not be weighted. In the illustrated example, for a pallet 810D, a rack opening 804B provides a highest optimization value (e.g., 1.0) among several candidate rack openings ding 804A-E), and can thus be ed as an optimal rack opening for the pallet 810D. [0089[ illustrates an example technology 900 for scaling a pallet storage area, such as one or more storage racks 902, depending on an occupancy of the area. For example, a computer system 904 can determine an occupancy rate of the storage racks 902 (Step A), and scale the storage racks 902 based on the occupancy rate (Step B). The occupancy rate can be determined to re?ect an occupancy at various times, such as at a predetermined past time, at a t time, or at a ermined future time. In some implementations, the occupancy rate can be determined or ted based on the current inventory, the historical inventory, and/or various other factors. [0090[ When an occupancy rate of the storage racks 902 is determined, an ive area 920 of the storage racks 902 can be adjusted based on the occupancy rate, thereby increasing the space utilization of the storage racks. By way of example, if the storage racks are ed to be occupied only 75% for a predetermined period of time in the future, the effective area 920 of the storage racks 902 can be reduced to a half of the storage racks 902, and the sections 906A-D of the storage racks can be scaled to the reduced effective area 920. The remaining area 930 of the storage racks 902 are unused. To maximize a labor cost saving, the effective area and the scaled sections of the storage racks 902 can be ed at the front part of the storage racks 902 close to the loading area 908. [0091[ is a block diagram of computing devices 1000, 1050 that may be used to implement the systems and methods described in this document, as either a client or as a server or plurality of servers. Computing device 1000 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, ames, and other appropriate computers. ing device 1050 is intended to represent various forms of mobile devices, such as personal digital assistants, cellular ones, smartphones, and other similar computing devices. The components shown here, their tions and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations described and/or claimed in this nt. [0092[ Computing device 1000 es a processor 1002, memory 1004, a storage device 1006, a high-speed interface 1008 ting to memory 1004 and high-speed expansion ports 1010, and a low speed interface 1012 connecting to low speed bus 1014 and e device 1006.

Each of the components 1002, 1004, 1006, 1008, 1010, and 1012, are interconnected using various busses, and may be mounted on a common motherboard or in other manners as appropriate. The processor 1002 can process instructions for execution within the computing device 1000, including instructions stored in the memory 1004 or on the storage device 1006 to display graphical information for a GUI on an external input/output device, such as display 1016 coupled to high-speed interface 1008. In other implementations, le processors and/or multiple buses may be used, as appropriate, along with multiple memories and types of memory.

Also, multiple computing devices 1000 may be connected, with each device ing portions of the necessary operations (e.g, as a server bank, a group of blade servers, or a multi-processor system). [0093[ The memory 1004 stores information within the computing device 1000. In one implementation, the memory 1004 is a volatile memory unit or units. In another implementation, the memory 1004 is a latile memory unit or units. The memory 1004 may also be another form of computer—readable medium, such as a magnetic or optical disk. [0094[ The storage device 1006 is capable of providing mass storage for the computing device 1000. In one implementation, the storage device 1006 may be or contain a computer- readable medium, such as a ?oppy disk device, a hard disk device, an optical disk device, or a tape device, a ?ash memory or other similar solid state memory device, or an array of devices, including devices in a storage area network or other con?gurations. A computer program product can be tangibly embodied in an information carrier. The computer program product may also contain ctions that, when executed, perform one or more methods, such as those described above. The information carrier is a computer— or machine-readable medium, such as the memory 1004, the storage device 1006, or memory on sor 1002. [0095[ The peed controller 1008 manages dth-intensive operations for the computing device 1000, while the low speed controller 1012 manages lower bandwidth-intensive operations. Such allocation of functions is an example only. In one implementation, the high- speed controller 1008 is coupled to memory 1004, display 1016 (e.g, through a graphics processor or accelerator), and to peed expansion ports 1010, which may accept various expansion cards (not shown). In the implementation, low-speed controller 1012 is coupled to storage device 1006 and low-speed expansion port 1014. The low-speed expansion port, which may include various communication ports (e.g., USB, Bluetooth, Ethernet, wireless Ethernet) may be coupled to one or more input/output devices, such as a keyboard, a ng device, a scanner, or a networking device such as a switch or router, e.g., through a network adapter. [0096[ The computing device 1000 may be implemented in a number of different forms, as shown in the ?gure. For example, it may be implemented as a rd server 1020, or multiple times in a group of such servers. It may also be implemented as part of a rack server system 1024. In addition, it may be implemented in a personal computer such as a laptop er 1022. Alternatively, components from computing device 1000 may be combined with other components in a mobile device (not shown), such as device 1050. Each of such devices may contain one or more of computing device 1000, 1050, and an entire system may be made up of multiple computing s 1000, 1050 communicating with each other. [0097 [ Computing device 1050 includes a processor 1052, memory 1064, an input/output device such as a display 1054, a communication interface 1066, and a transceiver 1068, among other components. The device 1050 may also be provided with a storage device, such as a microdrive or other device, to provide additional storage. Each of the ents 1050, 1052, 1064, 1054, 1066, and 1068, are onnected using various buses, and several of the ents may be mounted on a common motherboard or in other manners as appropriate. [0098[ The processor 1052 can execute ctions within the computing device 1050, including instructions stored in the memory 1064. The processor may be implemented as a chipset of chips that include separate and multiple analog and digital processors. Additionally, the processor may be implemented using any of a number of architectures. For example, the processor may be a CISC (Complex Instruction Set Computers) processor, a RISC (Reduced Instruction Set Computer) processor, or a MISC (Minimal Instruction Set Computer) processor.

The processor may provide, for example, for coordination of the other components of the device 1050, such as control of user aces, applications run by device 1050, and ss communication by device 1050. [0099[ Processor 1052 may communicate with a user through control interface 1058 and display interface 1056 coupled to a display 1054. The display 1054 may be, for example, a TFT (Thin-Film-Transistor Liquid Crystal Display) display or an OLED (Organic Light Emitting Diode) display, or other appropriate display technology. The display interface 1056 may se appropriate circuitry for driving the display 1054 to present graphical and other information to a user. The control interface 1058 may receive commands from a user and t them for submission to the processor 1052. In addition, an external interface 1062 may be provide in communication with processor 1052, so as to enable near area communication of device 1050 with other s. External ace 1062 may provided, for example, for wired communication in some implementations, or for wireless communication in other implementations, and multiple interfaces may also be used. [0100[ The memory 1064 stores information within the ing device 1050. The memory 1064 can be implemented as one or more of a computer-readable medium or media, a volatile memory unit or units, or a non-volatile memory unit or units. Expansion memory 1074 may also be provided and connected to device 1050 through expansion interface 1072, which may include, for example, a SHVIM (Single In Line Memory Module) card interface. Such expansion memory 1074 may provide extra storage space for device 1050, or may also store ations or other ation for device 1050. Specifically, expansion memory 1074 may include instructions to carry out or supplement the processes described above, and may include secure information also. Thus, for example, expansion memory 1074 may be provide as a security module for device 1050, and may be mmed with instructions that permit secure use of device 1050. In addition, secure applications may be provided via the SIMM cards, along with additional information, such as placing fying information on the SIMM card in a non- le manner. [0101[ The memory may include, for example, ?ash memory and/or NVRAM memory, as sed below. In one implementation, a computer program product is tangibly embodied in an information carrier. The computer program product contains instructions that, when executed, perform one or more methods, such as those described above. The ation carrier is a er— or machine-readable medium, such as the memory 1064, expansion memory 1074, or memory on processor 1052 that may be received, for example, over transceiver 1068 or external interface 1062. [0102[ Device 1050 may communicate wirelessly h communication interface 1066, which may e digital signal processing circuitry where necessary. Communication interface 1066 may provide for communications under various modes or protocols, such as GSM voice calls, SMS, EMS, or MIVIS messaging, CDMA, TDMA, PDC, WCDMA, 00, or GPRS, among others. Such communication may occur, for example, through radio-frequency transceiver 1068. In addition, short-range communication may occur, such as using a Bluetooth, WiFi, or other such transceiver (not shown). In addition, GPS (Global Positioning System) receiver module 1070 may e additional navigation- and location-related wireless data to device 1050, which may be used as appropriate by applications running on device 1050. [0103[ Device 1050 may also communicate y using audio codec 1060, which may receive spoken information from a user and convert it to usable digital information. Audio codec 1060 may likewise generate audible sound for a user, such as through a speaker, e.g., in a handset of device 1050. Such sound may include sound from voice one calls, may e recorded sound (e.g., voice messages, music files, etc.) and may also include sound generated by applications operating on device 1050. [0104[ The computing device 1050 may be implemented in a number of different forms, as shown in the ?gure. For example, it may be implemented as a cellular telephone 1080. It may also be implemented as part of a smartphone 1082, personal digital assistant, or other similar mobile device. [0105[ Additionally ing device 1000 or 1050 can include Universal Serial Bus (USB) ?ash drives. The USB ?ash drives may store operating s and other applications. The USB ?ash drives can include input/output components, such as a wireless itter or USB connector that may be inserted into a USB port of another computing device. [0106[ Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, specially designed ASICs (application speci?c integrated circuits), computer hardware, e, software, and/or combinations thereof. These various implementations can include implementation in one or more er ms that are able and/or retable on a programmable system including at least one programmable processor, which may be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device. [0107 [ These computer programs (also known as programs, software, software applications or code) include machine instructions for a programmable processor, and can be implemented in a high-level procedural and/or obj ect-oriented programming language, and/or in assembly/machine language. As used herein, the terms “machine-readable medium77 (L computer- readable medium” refers to any computer program product, apparatus and/or device (e.g., magnetic discs, l disks, , Programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine ctions as a machine-readable signal. The term “machine- readable signal” refers to any signal used to provide machine instructions and/or data to a programmable sor. [0108[ To e for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the er. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory ck, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input. [0109[ The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware ent (e.g., an application server), or that es a front end component (e.g., a client computer having a graphical user interface or a Web browser h which a user can interact WO 02111 2020/061235 with an implementation of the systems and techniques described here), or any ation of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area k (“LAN”), a wide area network (“WAN”), peer-to-peer networks (having ad-hoc or static members), ind computing infrastructures, and the Internet. [0110[ The computing system can include clients and servers. A client and server are generally remote from each other and typically ct through a communication network. The onship of client and server arises by virtue of computer programs running on the tive computers and having a -server onship to each other. [0111[ While this speci?cation contains many speci?c implementation details, these should not be construed as limitations on the scope of any inventions or of what may be claimed, but rather as descriptions of features speci?c to particular implementations of particular inventions.

Certain features that are described in this speci?cation in the context of separate implementations can also be ented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable sub-combination. er, although es may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a sub-combination or variation of a sub- combination. [0112[ Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all rated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous.

Moreover, the separation of various system components in the implementations described above should not be understood as requiring such tion in all implementations, and it should be understood that the described program components and systems can generally be ated together in a single software product or packaged into multiple software products.

W0 2021/‘102111 [0113[ Thus, particular implementations of the subject matter have been described. Other implementations are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve ble results. In addition, the processes depicted in the accompanying ?gures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain entations, multitasking and parallel processing may be advantageous.

Claims (20)

1. A system for managing a ity of pallets in a warehouse, the system comprising: a plurality of storage racks having a plurality of rack openings; a database that is programmed to store pallet allocation data that associates pallet storage durations with the plurality of rack openings; and a er system including one or more processors that are programmed to perform operations ing: determining an expected e duration of a target pallet in the warehouse; determining a height of the target pallet; calculating optimization values for the plurality of rack openings based on the expected storage duration and the height of the target pallet; identifying, among the plurality of rack openings, a target rack g as a storage location for the target pallet, the target rack opening having an optimization value that satisfies a preset requirement; and transmitting information identifying the storage location to equipment for placement of the target pallet, the equipment then utilizing the transmitted information to place the target pallet, wherein each of the optimization values includes a combination of a duration match value and a height match value for each of the plurality of rack gs, wherein the duration match value for a rack opening with respect to a pallet represents proximity in duration value between the pallet storage duration associated with the rack opening and the ed storage duration of the pallet, and wherein the height match value for the rack opening with t to the pallet represents proximity in measurement n a height of the rack opening and a height of the pallet.

2. The system of claim 1, wherein the optimization value of the target rack opening satisfies the preset requirement based on the optimization value exceeding a old value.

3. The system of claim 1, wherein the optimization value of the target rack opening satisfies the preset requirement based on the preset optimization value falling within a threshold range.

4. The system of claim 1, wherein determining the expected storage duration of the target pallet comprises: ining a duration percentile of the target pallet.

5. The system of claim 1, wherein the expected storage duration of the target pallet is determined based on historical inventory data, the historical inventory data identifying s stored in the warehouse, times at which the s were stored, and durations in which the pallets were stored.

6. The system of claim 1, wherein determining the expected storage duration of the target pallet comprises: predicting the expected e duration using a machine learning thm based on a plurality of input factors.

7. The system of claim 6, wherein the plurality of input factors include at least one of a type of items contained in the pallet, a customer of the pallet, a day of week, a week of year, a pallet description, an item description, and historical inventory data.

8. The system of claim 1, n the pallet storage durations of the plurality of rack openings are determined based on a distance between each of the plurality of rack openings and an entrance of the warehouse.

9. The system of claim 1, wherein the plurality of rack openings are mapped to different pallet duration percentiles, wherein the expected storage duration of the target pallet is identified as a first pallet duration percentile of the pallet duration percentiles.

10. The system of claim 1, wherein the plurality of rack openings includes a first rack opening and a second rack opening being arranged farther from an ce of the use than the first rack g, the first rack opening mapped to a first on percentile and the second rack opening mapped to a second duration percentile greater than the first duration percentile.

11. The system of claim 1, wherein the operations further e: determining one of the ity of rack openings based on a cost function of the expected storage duration and the height of the target pallet.

12. The system of claim 1, wherein the duration match value and the height match value are inversely weighted.

13. The system of claim 1, n the equipment comprises a forklift that es a user interface programmed to automatically output the information identifying the storage location in response to receiving the transmitted information.

14. The system of claim 1, wherein the plurality of storage racks includes one or more horizontal bars adjustable along a plurality of elevations on the storage racks to define the plurality of rack openings within the storage racks.

15. A computer-implemented method for managing a plurality of pallets in a warehouse, the method comprising: identifying, using at least one computing device, a target pallet red to the warehouse; accessing, using the at least one computing , a database that is programmed to store pallet allocation data that associates pallet storage durations with a plurality of rack openings in at least one storage rack; determining, using the at least one computing device, an expected storage duration of the target pallet in the warehouse; determining, using the at least one computing device, a height of the target ; calculating, using the at least one computing device, optimization values for the plurality of rack openings based on the expected storage duration and the height of the target pallet; identify, among the plurality of rack openings, a target rack opening as a e location for the target pallet, the target rack opening having an zation value that satisfies a preset requirement; and transmitting, using the at least one computing device, information identifying the storage on to equipment for placement of the target pallet, the ent then utilizing the transmitted information to place the target pallet, wherein each of the optimization values includes a combination of a duration match value and a height match value for each of the ity of rack openings, wherein the duration match value for a rack opening with respect to a pallet represents proximity in duration value between the pallet storage duration associated with the rack opening and the expected storage duration of the pallet, and wherein the height match value for the rack opening with respect to the pallet represents proximity in measurement n a height of the rack g and a height of the pallet.

16. The method of claim 15, wherein the optimization value of the target rack opening satisfies the preset requirement based on the optimization value exceeding a old value.

17. The method of claim 15, wherein the optimization value of the target rack opening satisfies the preset requirement based on the preset optimization value falling within a threshold range.

18. The method of claim 15, wherein determining the expected storage duration of the target pallet comprises: determining a duration percentile of the target pallet, wherein the expected storage duration of the target pallet is determined based on historical inventory data, the historical inventory data identifying pallets stored in the warehouse, times at which the s were stored, and durations in which the s were stored.

19. The method of claim 15, n determining the expected storage duration of the target pallet comprises: predicting the expected storage duration using a machine learning algorithm based on a plurality of input factors, wherein the plurality of input factors include at least one of a type of items contained in the pallet, a er of the pallet, a day of week, a week of year, a pallet description, an item description, and historical inventory data.

20. The method of claim 15, wherein the pallet storage durations of the plurality of rack openings are determined based on a distance between each of the plurality of rack gs and an entrance of the warehouse.